点击次数：

论文类型：会议论文

发表时间：2015-06-14

收录刊物：EI、CPCI-S、Scopus

卷号：9243

页面范围：150-158

关键字：Feature selection; Symmetrical uncertainty; Feature grouping; Genetic algorithm

摘要：Feature selection technique has shown its power in analyzing the high dimensional data and building the efficient learning models. This study proposes a feature selection method based on feature grouping and genetic algorithm (FS-FGGA) to get a discriminative feature subset and reduce the irrelevant and redundancy data. Firstly, it eliminates the irrelevant features using the symmetrical uncertainty between features and class labels. Then, it groups the features by Approximate Markov blanket. Finally, genetic algorithm is applied to search the optimal feature subset from the different groups. Experiments on the eight public datasets demonstrate the effectiveness and superiority of FS-FGGA in comparison with SVM-RFE and ECBGS in most cases.